ADULT PACIFIC LAMPREY MIGRATION IN THE LOWER COLUMBIA RIVER: 2010 RADIOTELEMETRY AND HALF-DUPLEX PIT TAG STUDIES

Transcription

1 Technical Report IDAHO COOPERATIVE FISH AND WILDLIFE RESEARCH UNIT ADULT PACIFIC LAMPREY MIGRATION IN THE LOWER COLUMBIA RIVER: 2010 RADIOTELEMETRY AND HALF-DUPLEX PIT TAG STUDIES A Report for Study Code ADS-P-00-8 by M. L. Keefer, C. C. Caudill, E. L. Johnson, T. S. Clabough, M. A. Jepson U.S. Geological Survey, Idaho Cooperative Fish and Wildlife Research Unit University of Idaho, Moscow, ID and M. L. Moser Northwest Fisheries Science Center, NOAA Fisheries 2725 Montlake Blvd. East, Seattle, WA For U.S. Army Corps of Engineers Portland District, Portland OR 2011

2 Technical Report ADULT PACIFIC LAMPREY MIGRATION IN THE LOWER COLUMBIA RIVER: 2010 RADIOTELEMETRY AND HALF DUPLEX PIT-TAG STUDIES A Report for Study Code ADS-P-00-8 by M. L. Keefer, C. C. Caudill, E. L. Johnson, T. S. Clabough, M. A. Jepson U.S. Geological Survey, Idaho Cooperative Fish and Wildlife Research Unit Department of Fish and Wildlife Resources University of Idaho, Moscow, ID and M. L. Moser Northwest Fisheries Science Center, NOAA Fisheries 2725 Montlake Blvd. East, Seattle, WA For U.S. Army Corps of Engineers Portland and Walla Walla Districts 2011 ii

3 Acknowledgements Many people assisted with the field work and data compilation for this report and its successful completion was made possible through their efforts. Steve Lee, Dan Joosten, Howard Pennington, Hilary Griffin, and Dennis Queampts were responsible for collecting and tagging lampreys. Brian Burke and Kinsey Frick helped manage the radiotelemetry data. Charles Boggs, Travis Dick, Mark Morasch, and Matt Knoff maintained and downloaded the monitoring equipment. We also thank the staff of Pacific States Marine Fisheries Commission s Kennewick Field Office, especially Darren Chase and Don Warf. The U.S. Army Corps of Engineers provided funding for this study; we thank Sean Tackley, David Clugston, Derek Fryer, Tammy Mackey, Jon Rerecich, Ben Hausmann, Kasey Welch, Miro Zyndol, Robert Cordie, Brad Eby, and Mark Plummer. iii

4 Table of Contents Executive Summary... v Introduction... 1 Methods... 1 Lamprey Collection and Tagging... 1 Monitoring Sites... 2 Data Analyses... 2 Results... 5 Lamprey Collection and Tagging... 5 Double-Tagged Lampreys... 6 Upstream Progression... 6 Dam-to-Dam Escapement... 7 Passage Times and Rates Diel Passage Last Detection Summary HD PIT-Tagged Lampreys (Excluding Double-Tagged Fish) Upstream Progression Dam-to-Dam Escapement Passage Times and Rates Last Detection Summary Detection of lampreys tagged in Efficiency of the Radiotelemetry Arrays at Dams Antenna Detection Efficiency Evaluations Using Double-Tagged Lampreys Detection Efficiency Discussion Escapement Passage Times Tagging and Handling Effects and Tag Performance Conclusions References Appendix A: Maps of Monitoring Sites in Appendix B: 2009 Columbia River Discharge and Temperature Profiles iv

5 Executive Summary We tagged adult Pacific lamprey (Lampetra tridentata / Entosphenus tridentatus) with half duplex (HD) passive integrated transponder (PIT) tags and/or radio transmitters and monitored their passage and migration behaviors at Bonneville, The Dalles, John Day, McNary, Ice Harbor, Lower Monumental, Lower Granite, and Priest Rapids dams. Our objectives were to calculate lamprey passage times, to estimate escapement past the monitored sites, and to evaluate potential physiological and environmental correlates with lamprey migration through the study area. We also examined telemetry system detection efficiencies. Near record-low adult lamprey counts at Bonneville Dam in 2010 resulted in reduced lamprey collection and necessitated modifications to our original study objectives. To remain below a 2% adult lamprey handling threshold (of the estimated total Bonneville count) established by regional managers, we prioritized radio tagging and eliminated study objectives associated with lampreys tagged only with HD-PIT tags. In total, we tagged 312 lampreys with both radio transmitters, and all of these also received HD-PIT tags. An additional 13 lampreys were tagged with HD-PIT tags only before objectives were re-prioritized. The 2010 escapement estimate from release below Bonneville Dam to top-of-ladder antennas was 41% for radio-tagged fish, the highest estimate among the radiotelemetry studies. Escapement from the top of Bonneville Dam to top-of-ladder antennas at The Dalles Dam (55%) and from the top of The Dalles Dam to the top of John Day Dam (49%) was also higher than in recent years. Large lampreys were significantly more likely than small lampreys to pass through most damto-dam reaches, and early migrants were more likely than late migrants to escape upstream. Lamprey migration times were highly variable, but tended to be slow at dams and relatively rapid through reservoirs. Median passage times for radio-tagged fish were 12.6 days from release to the top of Bonneville Dam, 4.1 days from tailrace to top-of-ladder antennas at The Dalles Dam, and 1.5 days between tailrace and top-of-ladder antennas at John Day Dam. Median reservoir passage times were 2.7 d (Bonneville) and 1.8 d (The Dalles). Lampreys generally migrated faster later in the summer through most reaches, coincident with increasing river temperatures and decreasing river discharge. Twenty-five percent of the lampreys that passed Bonneville Dam eventually entered tributaries, including the Snake River, or passed the uppermost monitoring sites at Priest Rapids Dam. The largest group in this category was recorded in the Deschutes River. Lampreys that returned to the Deschutes and John Day River were collected at Bonneville Dam relatively early in the run. In addition, most lampreys that returned to the upper basin were in the largest size classes, indicating probable size and migration timing effects on lamprey distribution in the basin. Almost all of the remaining 75% of the sample was last detected at main stem dams or in dam tailraces. At top-of-ladder sites where both radiotelemetry and HD-PIT detection systems were deployed, the radiotelemetry arrays were more efficient (total efficiency = 92%) than HD antennas (82%) at detecting double-tagged fish, although some fish were detected only on the HD-PIT system. The PIT tags were also detected at lamprey passage systems (LPS) systems at Bonneville Dam, improving overall detection efficiency. The combined basin-wide radiotelemetry results in 2010 indicated improved passage efficiency at Bonneville Dam and through some upstream reaches. Recent modifications made to fishway operations and structures likely improved lamprey passage at dams (especially Bonneville Dam). v

6 Introduction Pacific lamprey (Lampetra tridentata / Entosphenus tridentatus) is the largest lamprey species in the Columbia and Snake rivers. Pacific lampreys are anadromous, with parasitic adults spending 1-4 years in the ocean before returning to spawn in freshwater rivers (Beamish 1980; Close et al. 2002; Moser and Close 2003). Recent studies suggest that Pacific lamprey abundance has steadily declined in the Columbia River basin and in other regional rivers since the early 1960 s (Kostow 2002; Clemens et al. 2010). Habitat loss, river impoundment, ocean conditions, and water pollution have all likely contributed to the decline. Lampreys are also relatively poor swimmers and have difficulty passing through Columbia and Snake River dam fishways designed for adult salmonids (Moser et al. 2002b; Johnson et al. 2009a; Keefer et al. 2010a). Monitoring Columbia River basin lamprey populations has been a challenge. Lamprey counts at dam fish ladders can only be used as indicators of relative abundance and general run timing (e.g., Keefer et al. 2009b) because counts generally take place only during the day, most lamprey passage occurs at night, and counting facilities were not designed to accurately enumerate lampreys (Moser et al. 2002a; Robinson and Bayer 2005; Clabough et al. 2008). Radiotelemetry has been used in a series of studies over the last fourteen years to improve monitoring, identify problem passage areas, and estimate survival of adult Pacific lamprey in the basin (e.g., Moser et al. 2002b, 2005; Johnson et al. 2009a; Keefer et al. 2009a). Starting in 2005, half duplex (HD) passive integrated transponder (PIT) tag monitoring sites have been deployed at dams to monitor PIT-tagged adult lamprey. Like radio transmitters, PIT tags are uniquely identifiable, allowing monitoring of individual fish. PIT tags are also relatively small and inexpensive and are not limited by battery life, a useful feature given that some adult lamprey overwinter in the Columbia River main stem (Daigle et al. 2008) and some lampreys are too small for radio transmitters. HD-PIT tags were selected for Pacific lamprey passage evaluations to avoid potential tag collisions with the full-duplex (FDX) PIT tags used to monitor salmonids in the basin and because HD-PIT tags have longer read ranges. The objectives of the 2010 studies described in this report were to use both radiotelemetry and HD-PIT systems to: (1) calculate adult lamprey passage rates past multiple dams and reservoirs; (2) estimate lamprey escapement past multiple dams, through individual dam-to-dam reaches, and into tributaries; (3) examine potential physiological and environmental correlates with upstream passage; (4) compare results obtained using each tag type; and (5) to evaluate radiotelemetry and HD-PIT system detection efficiencies and radio tag effects using double-tagged fish. A description of finescale lamprey passage behaviors at individual dams and an evaluation of passage at the modified Cascades Island fishway will be presented in separate reports (Clabough et al. 2011a, b). Lamprey Collection and Tagging Methods Lampreys used in this study were collected at night in traps at Bonneville Dam (Columbia River kilometer [rkm] 235). Traps were located near the Adult Fish Facility and at the Washington-shore fishway entrance and additional fish were captured in portable traps. In 2010, 13 lampreys were tagged with half-duplex passive integrated transponder (HD-PIT) tags only, and 312 were tagged with radio transmitters and HD-PIT tags. No fish with girth < 9 cm at the dorsal fin were radio-tagged because past radiotelemetry studies have indicated increased tag effects for the smallest lampreys (Moser et al. 2007). All fish with only a HD-PIT tag were released approximately three kilometers downstream near Hamilton Island. Double-tagged fish were released near Hamilton Island (n = 154) or near Tanner Creek (n = 158) at Columbia River rkm

7 Before tagging, all fish were anaesthetized using 60 ppm (3 ml 50 L -1 ) clove oil measured (length and girth to the nearest mm), and weighed (nearest g). HD-PIT fish were then outfitted with a uniquely-coded, glass-encapsulated HD-PIT tag (Texas Instruments, 4 32 mm, 0.8 g). HD-PIT tags were surgically implanted in the body cavity of anaesthetized fish through a small incision (< 1 cm) along the ventral midline and in line with the anterior insertion of the first dorsal fin as described in Moser et al. (2006). Uniquely-coded radio tags (18.3 mm length, 8.3 mm diameter, 2.1 g in water; model NTC-4-2L, Lotek Wireless Inc.) were surgically implanted using the methods described in Moser et al. (2002a), with a PIT tag inserted through the same incision. An additional physiological measure, muscle lipid content (% lipid), was collected for all lampreys using a Distell fat meter (e.g., Crossin and Hinch 2005). Fat meter readings were converted to estimated % lipid (wet weight basis) using the regression equation percent lipid = * reading (P < 0.01; r 2 = ; n = 33). The regression equation was developed by comparing Fatmeter readings taken in previous years on live lamprey captured at Bonneville (n = 20) and McNary (n = 13) dams to lipid levels determined by biochemical proximate analysis on the same individuals following euthanasia. Proximate analysis was performed by the Wildlife Habitat Nutrition Laboratory, Washington State University, Pullman, Washington. Collection and tagging protocols were reviewed and approved by the University of Idaho Institutional Animal Care and Use Committee. Monitoring Sites Lamprey movements were monitored using an array of fixed-site radiotelemetry antennas and receivers (Table 1) and HD-PIT interrogation sites (Table 2). Aerial antennas were used to monitor dam tailraces and several tributary confluence areas. Underwater antennas and some directed aerial antennas detected radio-tagged fish as they approached, entered, and exited fishway openings, movements inside fishways and transition pools, and exits from ladders into dam forebays. Maps showing the positions of monitoring sites at dams are provided in Appendix A. Underwater HD-PIT antennas were located inside dam fishways at the four lower Columbia River dams, and at Priest Rapids and Ice Harbor dams. Antennas were located near top-of-ladder exits at all dams. At Bonneville Dam, additional sites were located at lamprey bypass structures (LPS), and inside the Washington-shore and Cascades Island fishway entrances. Antennas were also located near transition pools and/or the overflow weir portions of ladders at McNary and Ice Harbor dams and below the south (east) top-of-ladder site at The Dalles Dam (Table 2). Maps showing position of HD- PIT monitoring sites are provided in Appendix B. Data Analyses Reach escapement rates were calculated by dividing the number of lamprey known to pass an upstream HD or radiotelemetry site by those known to pass a site downstream or by the number released. Fish were treated as having passed a site if they were detected at the site or at a location further upstream. Escapement rates were calculated across all release dates. Lamprey sizes (length, weight, and girth) were compared for groups that passed through a reach and those that did not using general linear models (PROC GLM, SAS) and analysis of variance. Three lampreys were recaptured in the adult trap and these fish were transported and released into an LPS at Bonneville Dam. Where appropriate, these fish were excluded from analyses of escapement and passage times. 2

8 Table 1. Radiotelemetry antenna sites used to monitor lamprey passage at lower Columbia River dams, in reservoirs, and in tributaries in See Appendix A for maps of antenna sites at individual dams. Antenna(s) Site Location Type Number Bonneville Dam Tailrace Aerial 2 PH 1, South-shore entrance Underwater 3 PH 1, North-shore entrance Underwater 5 PH 1, A-Branch transition pool Underwater 3 PH 1, A- and B-Branch junction pool Underwater 3 PH 1, Bradford Island makeup water channel Underwater 3 PH 1, Bradford Island exit Underwater 1 B-Branch entrance Underwater 5 B-Branch transition pool Underwater 3 Cascades Island entrance Underwater 4 Cascades Island transition pool Underwater 4 Cascades Island makeup water channel Underwater 10 Cascades Island LPS Underwater 1 PH 2, South-shore entrances Underwater 7 PH 2, North-shore entrances Underwater 7 PH 2, WA-shore transition pool Underwater 6 PH 2, WA-shore ladder and turnpool Underwater 4 PH 2, WA-shore entrance lamprey trap Underwater 3 PH 2, WA-shore ladder / UMT channel pool Underwater 5 PH 2, WA-shore ladder makeup water channel Underwater 3 PH 2, WA-shore counting window Underwater 7 PH 2, WA-shore ladder exit Underwater 1 The Dalles Dam Tailrace Aerial 2 South spillway entrance Underwater 2 Powerhouse entrances Underwater 4 East ladder entrance Underwater 3 East ladder entrance 1 Aerial 2 East ladder transition pool Underwater 5 East ladder exit Underwater 1 North ladder entrance Underwater 1 North ladder entrance / transition pool Underwater 6 North ladder exit Underwater 1 North spillway area 1 Aerial 3 South spillway area 1 Aerial 1 John Day Dam Tailrace Aerial 2 South ladder exit Underwater 1 North ladder exit Underwater 3 McNary Dam Tailrace Aerial 2 South-shore entrance Underwater 4 South-shore transition pool / ladder Underwater 7 South ladder exit Underwater 5 North powerhouse entrance Underwater 3 North-shore entrance Underwater 3 North-shore transition pool / ladder Underwater 4 North ladder exit Underwater 4 3

9 Table 1. Continued. Antenna(s) Site Location Type Number Priest Rapids Dam East ladder exit Underwater 1 West ladder exit Underwater 1 Ice Harbor Dam Tailrace Aerial 2 South-shore entrance Underwater 3 South-shore transition pool / ladder Underwater 6 South ladder exit Underwater 2 North-shore entrance Underwater 2 North-shore transition pool / ladder Underwater 5 North ladder exit Underwater 3 Reservoir sites Fort Raines (rkm 235.3) Aerial 1 Bridge of the Gods (rkm 238.6) Aerial 1 Tributaries Willamette River Aerial 1 Eagle Creek Aerial 1 Herman Creek Aerial 1 Rock Creek Aerial 1 Wind River Aerial 1 Little White Salmon River Aerial 1 White Salmon River Aerial 1 Hood River Aerial 1 Klickitat River Aerial 1 Fifteenmile Creek Aerial 1 Deschutes River Aerial 1 John Day River Aerial 1 Yakima River Aerial 1 1 Data from aerial fishway antennas were not used for migration-scale analyses Lamprey migration times (d) and passage rates (km d -1 ) were calculated from release to top-ofladder HD-PIT and radiotelemetry antennas at dams and between monitored sites. Linear regression was used to evaluate relationships between log-transformed lamprey passage times and the following: length, girth, weight, release date or date fish entered upstream reaches, river discharge, and water temperature either on the release date or the date each fish passed top-ofladder sites at dams. Analyses using environmental data should be considered conservative as environmental conditions were constantly changing during the passage periods and the temporal variance may have obscured associations between behavior and environmental factors. Detection efficiencies for both HD-PIT and radiotelemetry sites were estimated by dividing the number of fish known to pass a site by the number that were detected at that site. These estimates were imprecise because fish could pass via unmonitored routes at many locations (e.g., navigation locks) and thus represent minimum estimates of detection efficiency. However, use of double-tagged fish allowed computation of more precise estimates of detection efficiencies. Detection efficiencies for both tag systems were evaluated at sites where radiotelemetry and HD-PIT antennas were in close proximity, primarily at top-of-ladder locations. 4

10 Table 2. Half-duplex PIT tag interrogation sites (antennas) used to monitor lamprey passage at lower Columbia and Snake river dams in See Appendix B for maps of antenna sites at individual dams. Site Location Number of antenna(s) Bonneville Dam PH 1, Bradford Island lamprey bypass 4 PH 1, Bradford Island exit 1 PH 2, WA-shore entrance 4 PH 2, WA-shore ladder 4 PH 2, WA-shore exit 1 PH 2, WA-shore lamprey bypass 2 Cascades Island 1 The Dalles Dam Below East ladder count window 4 East ladder exit (above count window) 4 North ladder exit 3 John Day Dam South ladder exit 1 North ladder exit 1 McNary Dam South-shore transition pool / ladder 4 South-shore exit 3 South-shore juvenile channel near exit 2 North-shore transition pool / ladder 4 North-shore exit 1 Ice Harbor Dam South-shore entrance 2 South-shore transition pool / ladder 4 South-shore exit 1 North-shore transition pool / ladder 4 North-shore exit 4 Priest Rapids Dam East ladder exit 3 West ladder exit 3 Lamprey Collection and Tagging Results The adult daytime lamprey count at Bonneville Dam in 2010 was the lowest recorded to date and this affected collection and tagging efforts. Specifically, the HD-PIT only tagging objective was dropped from the study plan early in the migration (Figure 1). Radio-tagging was made a priority and occurred from 30 May through 24 August, with the effort concentrated during traditional peak lamprey passage months of June and July. A total of 312 lampreys were radio-tagged, and all were also tagged with HD-PIT tags (i.e., they were all double-tagged). Thirteen lampreys were HD-PIT-tagged only, all in early June. The mean tag date for the double-tagged sample was 6 July. Lamprey size metrics and lipid content were positively correlated in the 2010 double-tagged sample (Figure 2, Table 3). Percent lipid was more variable than the other metrics, possibly in part because of the poor relationship between Fatmeter estimates and true lipid content (as measured by proximate analysis; see Methods). For example, for the double-tagged sample the coefficient of variation (CV) for mean %lipid was 30% versus 18% for weight, 7% for girth, and 6% for length (Table 3). In general, release date was weakly, but significantly, negatively correlated with lamprey length, girth, and weight (-0.27 < r < -0.18, P 0.05). 5

11 n = 6,234 counted n = 312 double-tagged Count at Bonneville n = 13 HDX-tagged only Tagged lampreys Apr 20 May 9 Jun 29 Jun 19 Jul 8 Aug 28 Aug 17 Sep 7 Oct Date Figure 1. Number of adult Pacific lamprey counted passing Bonneville Dam during the day and the numbers that were collected and double-tagged (top panel) or HD-PIT tagged only (bottom panel) in Table 3. Length, girth, weight, and percent lipid of adult Pacific lampreys that were collected and doubletagged with radio transmitters and HD-PIT tags at Bonneville Dam in Correlation coefficients (r) and significance values (P) for each pair of size metrics shown below. Length (cm) Girth (cm) Weight (g) Percent lipid (%) Type n Mean sd n Mean sd n Mean sd n Mean sd Double % 9.0% r P Length Girth <0.001 Length Weight <0.001 Length %Lipid <0.001 Girth Weight <0.001 Girth %Lipid <0.001 Weight %Lipid <0.001 Double-Tagged Lampreys Upstream Progression Of the 312 radio-tagged lamprey released downstream from Bonneville Dam, 276 (88.5%) were subsequently recorded at antennas at Bonneville Dam fishways or at sites further upstream (detections on aerial tailrace antennas near the release site were excluded for this estimate). Of these, three were recaptured in the adult trap and released directly into LPS systems as 6

12 part of LPS evaluations (two passed Bonneville and one did not). Whether the three would have passed the dam is unknown. Excluding the three recaptured lampreys, 40.5% of the released fish passed the dam (125/309). When the three recaptured lampreys were treated as passing, 41.0% of the released fish passed the dam (128/312). Dam passage estimates for fish detected at Bonneville antenna sites were 45.8% (125/273) excluding recaptured fish, and 46.4% (128/276) treating recaptured fish as passing. The most upstream recorded locations for the 184 fish that did not pass Bonneville Dam were: 36 (20%) recorded in the tailrace only; 39 (21%) approaching fishway entrances; 65 (35%) inside fishways, collection channels, or transition pools; and 44 (24%) in the ladders either upstream from transition pools or at sites near the tops of ladders. Fish that returned to the Bonneville fishways (i.e., were detected at fishway entrance antennas or further upstream) were larger than fish that did not return to the face of the dam. On average, returning fish were 2.5 cm longer, 43.9 g heavier, and had 0.4 cm wider girth (ANOVA P 0.015, length, weight, and girth) (Table 4). Returning fish were also tagged 17 d earlier, on average, than non-returning fish (P < 0.001). Lampreys that passed Bonneville Dam were also larger than fish that were recorded at fishway antennas but did not pass (length, weight, girth; P 0.022). The median tag date for all double-tagged lampreys was 6 July. Median passage dates at top-ofladder sites were 19 July at Bonneville Dam (n = 148 recorded), 23 July at The Dalles Dam (n = 68), 29 July at John Day Dam (n = 21), and 1 August at McNary Dam (n = 8). Top-of-ladder migration timing distributions for the double-tagged fish approximated those for daytime lamprey counts at the lower three dams (Figure 3). However, the very early portion of the run was under-represented at Bonneville Dam (as a result of downstream releases) and there was an under-representation of the later portion of the run at the upstream dams (as a result of tagging restrictions late in the run). The sample size was very small at McNary Dam. Dam-to-Dam Escapement Of 312 fish released, 40.4% (n = 126) were known to have passed Bonneville Dam and another 0.6% (n = 2) were recaptured in the adult trap, released into an LPS, and passed upstream. Treating the latter two fish as passing Bonneville Dam when calculating dam-todam escapement, 22.4% (n = 70) of 312 released passed The Dalles Dam, 10.9% (n = 34) passed John Day Dam, and 1.9% (n = 6) passed McNary Dam (Table 5). Escapement estimates from the top of Bonneville Dam (including the 2 recaptured fish) were 54.7% to the top of The Dalles Dam, 26.6% to the top of John Day Dam, and 4.7% to the top of McNary Dam. From the top of The Dalles Dam, estimates were 48.6% and 8.6% to the top of John Day and top of McNary dams, respectively (Table 5). In almost all cases, lampreys that passed dams upstream from Bonneville Dam and through damto-dam reaches were significantly larger than those that did not pass (Table 4). Length, weight, and girth differences were statistically significant (P < 0.05, ANOVA; no adjustment for multiple comparisons) for all reaches from the release site to the top of Bonneville, The Dalles, John Day, and McNary dams. Differences for the reaches between pairs of dams followed similar patterns, but statistical power was somewhat limited by small sample size. The %lipid metric did not differ significantly in any comparison. Tagging date was a significant effect in five of ten reaches; in all cases, successful fish were tagged earlier, on average (P < 0.05) (Table 4). 7

13 Weight (g) Girth (cm) Length (cm) 14 Length (cm) Percent lipid (%) Girth (cm) Length (cm) Weight (g) Percent lipid (%) Percent lipid (%) Weight (g) Girth (cm) Figure 2. Linear relationships between length, weight, girth, and %lipid metrics for adult lampreys doubletagged in 2010; two outliers excluded. See Table 3 for correlation coefficients. 8

14 Table 4. Mean double-tagged lamprey size metrics and tag dates in relation to their escapement through the monitored reaches in Top-of-ladder sites were used for the upper end of each reach. F and P values are from analysis of variance tests (ANOVA). Results include double-tagged fish that were recorded passing dams on the HD system only. Three fish that were recaptured at Bonneville Dam and released upstream were excluded from all reaches except the release-bonneville reach. Reach Passed Length (n) Weight (n) Girth (n) %Lipid (n) Tag date (n) Release-Bonneville 1 No 64.8 (36) (36) 10.9 (36) 29.3 (36) 22 Jul (36) Yes 67.3 (276) (275) 11.3 (275) 29.6 (276) 5 Jul (276) F P < <0.001 Bonneville 1 - Bonneville No 66.4 (147) (146) 11.2 (146) 28.8 (147) 6 Jul (147) top Yes 68.4 (126) (126) 11.4 (126) 30.6 (126) 3 Jul (126) F P < Release-Bonneville top No 66.1 (183) (182) 11.1 (182) 28.9 (183) 9 Jul (183) Yes 68.4 (126) (126) 11.4 (126) 30.6 (126) 3 Jul (126) F P <0.001 < Release-The Dalles top No 66.5 (241) (240) 11.1 (240) 29.4 (241) 10 Jul (241) Yes 69.0 (68) (68) 11.6 (68) 30.4 (68) 25 Jun (68) F P <0.001 <0.001 < <0.001 Release-John Day top No 66.6 (276) (275) 11.2 (275) 29.3 (276) 8Jul (276) Yes 70.4 (33) (33) 11.7 (33) 32.2 (33) 28 Jun (33) F P <0.001 <0.001 < Release-McNary top No 66.9 (303) (302) 11.2 (302) 29.5 (303) 7 Jul (303) Yes 73.0 (6) (6) 12.1 (6) 34.6 (6) 30 Jun (6) F P < Bonneville top-the No 67.7 (58) (58) 11.2 (58) 30.8 (58) 13 Jul (58) Dalles top Yes 69.0 (68) (68) 11.6 (68) 30.4 (68) 25 Jun (68) F P <0.001 The Dalles top-john No 67.7 (35) (35) 11.4 (35) 28.7 (35) 23 Jun (35) Day top Yes 70.4 (33) (33) 11.7 (33) 32.2 (33) 28 Jun (33) F P John Day top-mcnary No 69.9 (27) (27) 11.6 (27) 31.6 (27) 28 Jun (27) top Yes 73.0 (6) (6) 12.1 (6) 34.6 (6) 30 Jun (6) F P not including tailrace antennas 9

15 Bonneville The Dalles Lamprey counts at dams John Day Radio-tagged fish McNary Apr 20 May 9 Jun 29 Jun 19 Jul 8 Aug 28 Aug 17 Sep Date Figure 3. Daily numbers of adult Pacific lampreys counted passing dams via fish ladders (daytime only, gray lines) and the numbers of double-tagged lampreys that were detected at top-of-ladder radiotelemetry antennas (black bars) in

16 Table 5. Reach escapement estimates and antenna detection efficiencies for double-tagged adult lamprey in Two estimates are provided for each site or reach, one that includes both radiotelemetry (RT) and HD- PIT detections of double-tagged fish and a second based on radiotelemetry detections only. In addition, Bonneville top A includes three fish recaptured at the adult fish trap and released into LPS at Bonneville Dam; Bonneville top B excludes these fish from the release group. See Table 1 for antenna locations. Minimum Radiotelemetry Minimum All past site detections past site detections Site (all tag types) (%) (RT only) (%) Release Bonneville % % Bonneville top A N/A 123 N/A Bonneville top B % % The Dalles % % The Dalles top % % John Day % % John Day top % % McNary % % McNary top 2 (receiver outage) % % Reach escapement (%) All tag types Radiotelemetry only Difference Release-Bonneville % 88.1% 0.4% Release-Bonneville top A 41.0% 39.4% 1.6% Release-Bonneville top B 40.8% 39.2% 1.6% Release-The Dalles 33.3% 33.3% 0.0% Release-The Dalles top 22.4% 22.4% 0.0% Release-John Day 12.5% 12.2% 0.3% Release-John Day top 10.9% 10.3% 0.6% Release-McNary 3.5% 3.2% 0.3% Release-McNary top 1.9% 1.9% 0.0% Bonneville-Bonneville top A 46.4% 44.7% 1.7% Bonneville top A-The Dalles top 54.7% 56.9% -2.2% Bonneville top A-John Day top 26.6% 26.0% 0.6% Bonneville top A-McNary top 4.7% 4.9% -0.2% The Dalles top-john Day top 48.6% 45.7% 2.9% The Dalles top-mcnary top 8.6% 8.6% 0.0% John Day top-mcnary top 17.6% 18.8% -1.2% 1 all antennas, except tailrace 2 top-of-ladder and LPS antennas 3 all antennas, including tailrace Passage Times and Rates Median double-tagged lamprey passage times were d from the release site to the top of Bonneville Dam, 7.15 d between Bonneville and The Dalles dams, and 3.38 d between The Dalles and John Day dams (top-of-ladder sites at all dams, Table 6). Receiver outages at the top of McNary Dam precluded accurate passage time estimation in Median passage rates (upstream migration distance divided by passage time in days) in these reaches were < 1 km d -1 (release-bonneville top), 10.2 km d -1 (Bonneville-The Dalles, and 11.5 km d -1 (The Dalles- John Day). Passage rates through reservoirs only (i.e., from ladder exit at the downstream dam to upstream dam tailrace) were 25.7 km d -1 in the Bonneville reservoir and 20.9 km d -1 in The Dalles reservoir. 11

17 Median lamprey passage times through multiple dam-and-reservoir reaches were d (2.6 km d -1 ) from release to the top of The Dalles Dam, and d (3.7 km d -1 ) from release to the top of John Day Dam (Table 6). Table 6. Summary of double-tagged adult lamprey passage times through dam-to-dam and multi-dam reaches of the lower Columbia River in Fish recaptured at Bonneville Dam were excluded from reaches starting at release and ending at sites upstream from Bonneville Dam. Passage time (d) Reach n Median Mean Quartile 1 Quartile 3 Release to approach Bonneville fishway Release to enter Bonneville fishway Release to pass Bonneville Dam Release to The Dalles tailrace Release to pass The Dalles Dam Release to John Day tailrace Release to pass John Day Dam Release to McNary tailrace Release to pass McNary Dam Bonneville top to The Dalles tailrace Bonneville top to pass The Dalles Dam Bonneville top to John Day tailrace Bonneville top to pass John Day Dam Bonneville top to McNary tailrace Bonneville top to pass McNary Dam The Dalles tailrace to pass The Dalles Dam The Dalles top to John Day tailrace The Dalles top to pass John Day Dam The Dalles top to McNary tailrace The Dalles top to pass McNary Dam John Day tailrace to pass John Day Dam John Day top to McNary tailrace John Day top to pass McNary Dam McNary tailrace to pass McNary Dam In general, lamprey passage times decreased through the season as discharge decreased and temperature increased. An exception was that migration times from release to first approach at Bonneville Dam fishway was significantly, positively correlated (P < 0.05) with release date and water temperature, and negatively correlated with river discharge (Table 7). Passage times from release to exit into the Bonneville Dam forebay were positively correlated with lamprey size metrics and river discharge (0.25 r 0.30, P < 0.05) and negatively correlated with release date and water temperature (r = -0.35, P < 0.05). Passage times from release to sites at dams upstream from Bonneville Dam were almost all significantly (P < 0.05), positively correlated with river discharge and negatively correlated with release date and water temperature. Correlations with lamprey size metrics over the longer reaches were also positive, and most were significant (Table 8). The strong intercorrelations among flow, photoperiod, and water temperature make it difficult to attribute the observed associations to specific underlying mechanism(s) beyond the aggregate effect of season on migration rate. 12

18 Table 7. Correlation coefficients (r) for log-transformed radio-tagged lamprey passage times (d) in Predictor variables included size metrics (length, weight, girth) recorded at the time of tagging and date, Columbia River flow and temperature on the date fish were released or passed the top-of-ladder sites for each reach. Gray shading indicates P < Fish recaptured at Bonneville Dam were excluded from reaches starting at release and ending at sites upstream from Bonneville Dam. Correlation coefficient (r) Reach start Reach end n Length Weight Girth Date Flow Temp. Release Bonneville approach Release Bonneville entry Release Bonneville top Release The Dalles tailrace Release The Dalles top Release John Day tailrace Release John Day top Release McNary tailrace Release McNary top Bonneville top The Dalles tailrace Bonneville top The Dalles top Bonneville top John Day tailrace Bonneville top John Day top The Dalles top John Day tailrace The Dalles top John Day top Diel Passage Lamprey passage distributions at top-of-ladder sites clearly indicate that most passage occurred at night (Figure 4). This pattern was consistent at the four lower Columbia River dams (n = 1 at McNary Dam), with the majority of passage events between sunset and sunrise. However, some fish passed during almost all hours of the day and night. Daylight passage most often occurred between 0500 and 1000, suggesting that some fish that initiated dam passage at night continued migrating in the morning. Last Detection Summary About one third (33.7%) of the 312 double-tagged lampreys were last recorded at Bonneville tailrace antennas and another 20.8% were last recorded at Bonneville fishway antennas, including those outside the fishway entrances and at top-of-ladder exits (Table 8). Twentythree fish (7.4%) were last recorded downstream from the release sites, and one (0.3%) was mobiletracked in the Willamette River. Upstream from Bonneville Dam, six fish (1.9%) were recorded in Bonneville reservoir tributaries, 10 (3.2%) were last recorded at Bonneville forebay antennas, and 34 (10.9%) were last recorded in the tailrace of fishways of The Dalles Dam. In total, 51 fish (16.3%) were last recorded in the main stem between the Bonneville Dam exit and The Dalles Dam fishway. Another 11 fish (3.5%) were last recorded at the top-of-ladder exits at The Dalles Dam and 17 (5.4%) were last recorded in the Deschutes River. Twenty-one (6.7%) were at John Day Dam, eight (2.6%) were at McNary Dam, and one each (0.3%) was at Priest Rapids Dam, Lower Monumental Dam, and Lower Granite Dam (all tailrace, fishway, and ladder sites combined, Table 8). 13

19 30 Number of lamprey (n) Bonneville The Dalles John Day McNary Time of day Figure 4. Distributions of the times (binned by hour) that radio-tagged lamprey were detected passing top-ofladder antennas at Bonneville, The Dalles, John Day, and McNary dams in Table 8. Last recorded locations for 312 double-tagged adult Pacific lampreys in Site n % Site n % Willamette River 1 0.3% Deschutes River % Columbia R. below release site % John Day tailrace % At John Day Dam 1 0.3% Bonneville tailrace % John Day ladder exits % At Bonneville Dam % Bonneville ladder exits 7 2.2% John Day River 8 2.6% Bonneville forebay FTR 1 BOG % At McNary Dam 5 1.6% Klickitat River 4 1.3% McNary Dam ladder exits 3 1.0% Fifteen Mile Creek 2 0.6% Priest Rapids ladder exits 3 0.3% The Dalles tailrace % At The Dalles Dam % Lower Monumental ladder exit 1 0.3% The Dalles ladder exits % Lower Granite ladder exit 1 0.3% 1 Fort Raines; 2 Bridge of the Gods When lampreys were grouped based on final recorded locations, median release dates differed substantially among groups (Figure 5). On median, the earliest-timed fish were those that were last recorded at the top of The Dalles Dam (9 June), in the Deschutes River (14 June), in the John Day River (20 June), and at McNary Dam (23 June). The latest-timed groups were last recorded in the Bonneville forebay (26 July), in The Dalles tailrace (20 July), in the Klickitat River (16 July), and at or inside a Bonneville Fishway (15 July). 14

20 There were also clear among-fate differences in lamprey size (Figure 6). On median, lampreys were largest in the above McNary (576 g), McNary tailrace (560 g), John Day River (551 g), and top of John Day (541 g) groups. Lampreys were smallest in the groups last recorded in the Bonneville forebay (440 g), Bonneville tailrace (459 g), in the combined Klickitat River and Fifteen Mile Creek (482 g), and below the release site (486 g). We used multiple logistic regression models to help visualize the relationships between migration timing, lamprey size, reach escapement, and the various fate groups (Figure 7). In each model the date of reach entry (or release date) and lamprey weight were used as predictor variables and passage through the reach was the dependent variable. Larger lampreys were more likely (P 0.07) to pass through five of the six study reaches tested; the exception was The Dalles tailrace/fishway- The Dalles top. Early-timed lampreys were significantly (P < 0.05) more likely to pass through the release-bonneville approach, The Dalles tailrace/fishway-the Dalles top, and The Dalles top-john Day tailrace/fishway reaches. Final fate groups superimposed on the probability estimates showed some clear separation-by-date and separation-by-size patterns (Figure 7). All Lamprey n = 312 MCN or higher MCN tailrace John Day R JDD top JDD tailrace Deschutes R TDD top TDD tailrace Klickitat/15 Mile BON top / forebay BON tailrace Below release May 1 Jun 15 Jun 1 Jul 15 Jul 1 Aug 15 Aug 1 Sept Release date Figure 5. Distributions of double-tagged lamprey release dates downstream from Bonneville Dam by the final recorded locations for each fish. Tailrace locations include fish last recorded at fishway entrances or inside fishways without passing. Box plots show 5 th, 25 th, 50 th, 75 th and 95 th percentiles. 15

21 MCN or higher MCN tailrace All at MCN John Day R JDD top All past JDD JDD tailrace All at JDD Deschutes R TDD top All past TDD TDD tailrace All at TDD Klickitat/15 Mile BON top / forebay All past BON BON tailrace Below release All Lamprey Weight (g) Figure 6. Distributions of double-tagged lamprey weights (g) for adults entering each reach (white boxes) and by the final recorded locations for each fish (grey boxes). Tailrace locations include fish last recorded at fishway entrances or inside fishways without passing. Box plots show 5 th, 25 th, 50 th, 75 th and 95 th percentiles. HD PIT-Tagged Lampreys (Excluding Double-Tagged Fish) Upstream Progression Of the 13 lampreys released downstream from Bonneville Dam with only HD-PIT tags, 9 (69.2%) were subsequently recorded at one or more Bonneville Dam HD antennas inside fishways, at LPS systems, or at dams further upstream. A total of 7 fish passed Bonneville Dam (53.8% of the 13 released, and 77.8% of the 9 detected at one or more HD antennas) and one additional fish was recaptured and then released into an LPS system. The median tag date for HD-PIT tagged lampreys was 8 June about one month earlier than the double-tagged fish. This early tagging bias resulted directly from the early-season change in research objectives and precludes direct comparison with results from double-tagged fish or HD-PIT only fish in prior years. Median recorded passage dates at top-of-ladder sites were 25 June at Bonneville Dam (n = 8, including the recaptured fish), 5 July at The Dalles Dam (n = 3), and 19 July at John Day Dam (n = 2). Dam-to-Dam Escapement Of 13 fish released, 61.5% (n = 8) were known to have passed Bonneville Dam, 23.1 % (n = 3) passed The Dalles Dam, and 15.4% (n = 2) passed John Day Dam. Escapement from the top of Bonneville Dam was 37.5% to the top of The Dalles Dam and 25.0% to the top of John Day Dam. Escapement between ladder tops at The Dalles Dam and John Day dams was 66.7%. 16

22 Approach Bonneville Pass Bonneville 600 Weight Release date Approach The Dalles Bonneville fishway approach date Pass The Dalles 600 Weight Bonneville ladder exit date Approach John Day The Dalles tailrace / fishway date Pass John Day 600 Weight The Dalles ladder exit date John Day tailrace / fishway date Figure 7. Bubble plots showing the predicted probability of double-tagged lamprey passage through six study reaches (Release Approach Bonneville fishway; Approach Bonneville fishway pass Bonneville; top of Bonneville The Dalles tailrace/fishway; The Dalles tailrace/fishway pass The Dalles; top of The Dalles John Day tailrace/fishway; John Day tailrace/fishway pass John Day). Probabilities were calculated using the logistic regression model: passage (0,1) = lamprey weight + reach entry date. Larger bubbles reflect higher probability. Colored symbols are superimposed on model results and represent groups of lampreys with different final fate: blue = Klickitat River and Fifteen Mile Creek; red = Deschutes River; green = John Day River; black = at or above McNary Dam. Date 60 = 30 May, Date 200 = 17 October. 17

23 Passage Times and Rates Median HD-PIT tagged lamprey passage times were 16.6 d (n = 7) from the release site to the top of Bonneville Dam, 4.0 d (n = 3) between Bonneville and The Dalles dams, and 3.6 d (n = 2) between The Dalles and John Day dams, and 9.8 d between John Day and McNary dams (top-of-ladder sites at all dams). Median passage rates in these reaches were < 1 km d -1 (release-bonneville top), 18.3 km d -1 (Bonneville-The Dalles), and 10.8 km d -1 (The Dalles- John Day). As with the double-tagged data, the majority of dam passage events were between sunrise and sunset. Last Detection Summary About a third (30.8%) of the 13 lampreys with only HD-PIT tags were not detected after release near Hamilton Island. Another 46.1% were last recorded at HD antennas at Bonneville Dam, including one at the Cascades Island fishway entrance and five at top-of-ladder exit sites. One fish each (7.7%) were last recorded at top-of-ladder exits at The Dalles and John Day dams and one (7.7%) was last recorded inside a McNary fishway. Detections of lampreys tagged in 2009 A total of four lampreys tagged in 2009 were detected on HD antennas in All four were detected at Bonneville Dam between 5 June and 5 July, and three were detected at The Dalles Dam from 29 June to 9 July. Three of the four fish had been double-tagged in 2009 (1% of 300 released) and one was HD PIT-only (0.3% of 368 released). Efficiency of the Radiotelemetry Arrays at Dams Efficiencies described in this section were calculated by dividing the number of lampreys detected at a radiotelemetry array (multiple antennas) by the number known to have passed the array. (No HD records were included: see the Evaluations Using Double-Tagged Lampreys section for efficiencies calculated using double-tagged fish at individual sites). Overall, radio-tagged fish known to have reached Bonneville, The Dalles, and John Day dams were highly likely to be detected on one or more antennas (Table 5). At Bonneville Dam, 312 fish (100% of those released) were recorded at one or more sites. Of these, 306 (98.1%) were detected on aerial tailrace antennas (near the release sites). A total of 275 fish were recorded at antennas upstream from the Bonneville Dam tailrace (including outside fishways) and 237 were recorded inside Bonneville Dam fishways. Of the 237 recorded inside fishways, 207 (87.3%) were recorded passing an entrance antenna (many of the remaining 12.7% entered at unmonitored locations rather than passed undetected). A minimum of 126 lampreys were known to have volitionally passed Bonneville Dam via top-of-ladder and LPS routes, and 119 (94.4%) of these were detected as they passed arrays at top-of-ladder sites. Of the seven fish known to have passed Bonneville Dam that were not recorded at the exits, 4 (57%) were last detected at antennas in the auxiliary channel near the LPS at the top of the Bradford Island ladder; these fish likely passed the dam via the LPS. Another two fish (29%) probably passed the Washington-shore ladder during a receiver power outage on 13 July. Thus, detection efficiency at the top of the fishway sites during times of receiver operation was likely > 95%. A total of 104 fish were detected at The Dalles Dam antenna array. Fifty-five of the 104 (52.9%) were recorded on aerial tailrace antennas. A total of 93 fish were recorded at fishway entrances or inside fishways. Eighty-four were recorded inside fishways, 75 of which (89.2%) were detected by fishway entrance antennas. Seventy fish were known to pass The Dalles Dam, and all (100%) were recorded at top-of-ladder antennas (Table 5). Thirty-nine lampreys were recorded at John Day Dam antenna arrays, with 30 (76.9%) recorded on aerial tailrace antennas (two only after dam passage and subsequent fallback). Thirty-four fish were known to pass John Day Dam, and 32 (94.1%) of these were recorded at top-of-ladder antennas (Table 5). Two fish were detected only at HD-PIT antennas at the top of John Day ladders. Eleven 18

24 fish were recorded at McNary Dam. None (0%) were recorded at aerial tailrace antennas. Eight fish were recorded inside McNary fishways, of which 2 (18%) were recorded at fishway entrance antennas. Due to receiver outages at the Oregon Shore Fishway at McNary dam (where most adult lamprey pass), only one of six (17%) lampreys that likely passed the dam was recorded at the top-ofladder antenna arrays. Antenna Detection Efficiency Evaluations Using Double-Tagged Lampreys Detection Efficiency We calculated detection efficiencies for both radio and HD-PIT antennas at top-of-ladder sites at all four lower Columbia River dams using the double-tagged lampreys (Table 9). These estimates represent the percentage of fish that passed an antenna that was detected. The combined detection efficiencies for all top-of-ladder sites were 92.3% for the radiotelemetry sites and 81.2% of the HD sites. Detection efficiencies at individual radiotelemetry sites ranged from 93.3% at the John Day north ladder to 100% at several ladder sites, with the exception of McNary Dam, where outages at radiotelemetry sites reduced efficiencies. Most missed passages for the radiotelemetry antennas could be associated with power outages or damaged equipment. The HD-PIT detection efficiencies at individual sites ranged from 39.4% at The Dalles east ladder to 100% at both John Day ladders (Table 9). The low efficiency at The Dalles east was related to the size of the opening. Most of the fish missed at this top-of-ladder HD antenna site were recorded at the HD antenna installed downstream from the ladder top near the count window (Appendix A). The latter HD site did not have a radiotelemetry equivalent. Table 9. Detection efficiencies at top-of-ladder antennas calculated using double-tagged lampreys in Includes fish detected on at least one monitoring system. Detection efficiency Dam Ladder Total detected Radiotelemetry HD-PIT Bonneville South (Bradford) % 95.7% North (WA-shore) % 79.1% The Dalles South (East) % 39.4% North % 100.0% John Day South % 100.0% North % 100.0% McNary South 5 0.0% 20.0% North % 100.0% All dams All ladders % 82.2% 1 Does not include fish that passed through the makeup water channel because there was no paired PIT antenna Discussion The 2010 adult lamprey studies had multiple objectives that addressed lamprey migration in the Columbia River Hydrosystem at a variety of scales and using multiple technologies. The results summarized in this report primarily address reach-scale and system-wide migration using radiotelemetry. HD-PIT tag objectives became secondary in 2010 when small run size precluded tagging fish with PIT tags only. Companion 2010 study reports provide results for fine-scale 19